位置：成果数据库 > 期刊 > 期刊详情页
基于分裂EM算法的GMM参数估计
  • ISSN号：1002-8331
  • 期刊名称：《计算机工程与应用》
  • 时间：0
  • 分类：TP391[自动化与计算机技术—计算机应用技术;自动化与计算机技术—计算机科学与技术]
  • 作者机构：[1]安徽大学电子与信息系,合肥230031, [2]安徽农业大学计算机信息学院,合肥230036, [3]合肥工业大学计算机与信息学院,合肥230009
  • 相关基金：国家自然科学基金（No.30800663,No.U1135003）.
作者： 钟金琴[1,3], 辜丽川[2], 檀结庆[3], 李莹莹[3]
关键词： 高斯混合模型, 期望最大化, 参数估计, 模式分类, Gaussian Mixture Model （GMM）, Expectation Maximization （EM）, parameters estimation, pattern classification
中文摘要：

期望最大化（Expectation Maximization，EM）算法是一种求参数极大似然估计的迭代算法，常用来估计混合密度分布模型的参数。EM算法的主要问题是参数初始化依赖于先验知识且在迭代过程中容易收敛到局部极大值。提出一种新的基于分裂EM算法的GMM参数估计算法，该方法从一个确定的单高斯分布开始，在EM优化过程中逐渐分裂并估计混合分布的参数，解决了参数迭代收敛到局部极值问题。大量的实验表明，与现有的其他参数估计算法相比，算法具有较好的运算效率和估算准确性。

英文摘要：

The expectation maximization algorithm has been classically used to find the maximum likelihood estimates of parameters in mixture probabilistic models. Problems of the EM algorithm are that parameters initialization depends on some prior knowledge, and it is easy to converge to a local maximum in the iteration process. In this paper, a new method of estimating the parameter of GMM based on split EM is proposed, it starts from a single mixture component, sequentially split and estimates the parameter of the mixture components during expectation maximization steps. Extensive experiments show the advantages and efficiency of the proposed method.

同期刊论文项目
知识网格环境下基于动态数据挖掘与案例推理的农作物生物灾害预决策方法
期刊论文 28 会议论文 9
基于数字家庭的高清媒体内容分析理解中的非线性理论及关键技术
期刊论文 99 会议论文 26 获奖 1
同项目期刊论文
Optimized image retargeting using aesthetic-based cropping and scaling
A piecewise Pade-type rational interpolation of hybrid order in the interval [0,1]
Explicit determinant formulas of generalized restrictive Pade approximation
A Novel Image Decomposition Approach and its Applications
Robustness of orthogonal matching pursuit under restricted isometry property
A novel video denoising method based on total variationand recursive temporal filtering
Dual Generalized B-spline Bases and Their Applications in Approximation Problems
A new rationalcubic trigonometric Bezier curve with four shape parameters
Compression and reconstruction of time series data based on vector valued continued fraction
A novel depth map generation method and image quality test method of stereo television
A combined approximating and interpolating subdivision scheme with C2 continuity
A binary five-point relaxation subdivision scheme
A new extension of quadratic Bézier curves with multiple shape parameters
Perspective projection based on quaternion interpolation
Multiple degree reduction of Wang-Ball curves by using dual bases polynomials
Optimal Scaling Factor Assignment for Patchwise Image Retargeting
A blending method based on partial differential equations for image denoising
Multiple instance learning tracking method with local sparse representation
A fast mesh parameterization algorithm based on 4-point interpolatory subdivisio
Weighted rational quartic spline interpolation
A new variant of Newton’s method based on equidistant nodes
Local Texture Suppression Filter Based on Joint Degenerative Scheme.
Convergence of accelerated modulus based matrixsplitting iteration methods for linear complementarit
A Non-stationary Binary Three-point Approximating Subdivision Scheme
The iteration method for sample-based polynomial approximation ofrational B-spline curves
BAP: Bimodal attribute prediction for zero-shot image categorization
Irregular bivariate blending continued fractions interpolation in image resizing
An automatic video scratch removal based on Thiele type continued fraction
Video image restoration based on sparse representation
Real-time background subtraction based video surveillance of people by Integrating local texture pat
An adaptive image denoising model based on Tikhonov and TV regularizations
Two adaptive image pre-processing chains for face recognition rate enhancement,
Adaptive principal component analysis based wavelet transform and image de-noising for face recognit
A novel method of edge detection based on the Isotropic diffusion model and total variation model
New development of non-rigid registration
A new extension of quadratic Bezier curves with multiple shape parameters
A fast and efficient mesh segmentation method based on improved region growing
A novel algorithm for removal of salt and pepper noise using continued fractions interpolation
Hierarchical Ensemble of Background Models forPTZ-based Video Surveillance
Single Image Super-ResolutionUsing Maximizing Self-Similarity Prior
A fast and efficient mesh segmentation method based on improved regiongrowing
Perspective projection based on quaternioninterpolation
Acceleratedmodulus-based matrix splitting iteration methods for linear complementarity problem
Multi-targets tracking based on bipartitegraph matching
A method of image magnificationbased on radial basis function
Compact descriptor for local featureusing dominating center-symmetric local binary pattern
Super-resolution by polar Newton-Thiele's rational kernel in centralized sparsity paradigm
凸多边形域上的参数曲面自由变形
Automatic detection and removal of high-density impulse noises
Multi-scale patch-based sparse appearance model for robust object tracking
A method of Neville-like vector-valued blending rational interpolants based on continued fractions
A novel model of image segmentation based on watershed algorithm
Bi-cubic C-Coons surface and its applications
Bézier-like curves based on algebraic and exponential polynomials and their connection co
A class of generalized B-spline quaternion curves
Video noise reduction method using adaptive spatial -temporalfiltering
Image denoising with patch estimation and low patch-rank regularization
Least square geometric iterative fitting method forgeneralized B-spline curves with two different ki
An amalgammethod based on anisotropic diffusion and temporal filtering for videodenoising
Real-time Visual ObjectTracking via CamShift-Based Robust Framework
精确局部特征描述的表情识别
带平台伸缩函数的参数曲线变形
三角域上Said-Ball基的推广渐近迭代逼近
基于双边滤波改进的Harris特征点检测
四步八阶迭代法解非线性方程组
A Unified Ternary Curve Subdivision Scheme
一种改进的整体变分图像修复方法
多形状参数的指数均匀B样条曲线曲面
基于不光滑边界的变系数抛物型方程的高精度紧格式
三步五阶迭代方法解非线性方程组
一种基于Chebyshev迭代解非线性方程组的方法
一类新的（2n-1）点二重动态逼近细分
带互异权值的渐进迭代逼近算法及其应用
基于梯度Gabor直方图特征的表情识别方法
基于插值细分的逼近细分法
广义B样条曲线的节点去除与光顺算法
全局各向异性四边形主导网格重建方法
带一个形状参数的有理三次三角Bézier曲线
基于二阶矩的SIFT特征匹配算法
一种基于质量云的软件质量评估方法
Study of predictive method based on SVM optimal model selection
A novel simulation model to measure the performance of network traffic utilization in heterogeneous
基于XML的农作物病虫害数据转换研究与实现
本体演化研究进展
基于XML的农作物病虫害数据 转换研究与实现
基于用户检索和加权技术的领域本体演化研究
一种基于支持向量回归和动态特征选择的梨黑星病预测方法
语义Web应用研究综述
概念格理论与方法研究
基于敏感度分析的案例特征项权重算法的改进
基于茶虫害本体的智能检索系统设计
茶学本体学习中的概念抽取
基于语义Web服务的电子政务模型研究
属性重要性评分方法的改进
基于本体的水稻病害诊断专家系统研究
面向茶学领域本体的概念自动提取方法研究
一种基于核矩阵迭代学习的范例相似度算法
设计模式的一种形式化描述方法
一种基于支持向量回归的梨黑星病预测方法
DICOM查询/检索服务类SCP的设计与实现
基于本体推理的水稻虫害诊断系统研究
期刊信息
  • 《计算机工程与应用》
  • 北大核心期刊（2014版）
  • 主管单位:中国电子科技集团公司
  • 主办单位:华北计算技术研究所
  • 主编：怀进鹏
  • 地址：北京市海淀区北四环中路211号北京619信箱26分箱
  • 邮编：100083
  • 邮箱：ceaj@vip.163.com
  • 电话：
  • 国际标准刊号：ISSN：1002-8331
  • 国内统一刊号：ISSN：11-2127/TP
  • 邮发代号:82-605
  • 获奖情况:
  • 1. 2012年首批获得中国学术文献评价中心发布的 “...,2. 2001年获得新闻出版署“中国期刊方阵双效期刊”,3. 2008年首批入选国家科技部“中国精品科技期刊...,4.2003年-2011年连续获得工业和信息化部期刊最高...
  • 国内外数据库收录:
  • 俄罗斯文摘杂志,波兰哥白尼索引,美国剑桥科学文摘,英国科学文摘数据库,日本日本科学技术振兴机构数据库,中国中国科技核心期刊,中国北大核心期刊（2004版）,中国北大核心期刊（2008版）,中国北大核心期刊（2014版）,中国北大核心期刊（2000版）
  • 被引量:97887